Fundamental Philosophical Questions
Notes on Utilitarianism
Utilitarianism: The term was invented by Jeremy Bentham (early 19 century) to
describe his view that an action is right if it leads to the greatest overall happiness relative
to available alternatives. He summed up his view with the slogan “the greatest happiness
of the greatest number”. The expression is ambiguous, since in some circumstances one
could create greater happiness but for a smaller number, or less happiness for a greater
number. But it is clear that he favoured greatest (or maximum) aggregate happiness,
calculated by adding up all the happinesses of each individual (minus their pains). So the
“number” of individuals is only relevant as a factor in the overall aggregate, but has no
Bentham was also a hedonic utilitarian, meaning that he then defined happiness in
simple terms of pleasure minus pain. But he also used the general term ‘utility’(hence the
concept of utilitarianism) and spoke of ‘utils’as the units by which one calculated how
much good was produced by some action.
Other, more sophisticated versions, try to maximise not just pleasure (and minimise pain)
but to maximise welfare, well-being, or (especially popular among economists)
preference satisfaction. The latter is well-suited to a market economy, the appeal of
which stems from its responsiveness to the preferences of consumers.
Akey element of utilitarianism, and also itsAchilles’heel: it judges actions by whether
they maximise aggregate good in society, and therefore appears indifferent to the
distribution of the good. You add up all the pleasure, or happiness, ‘utils’, preferences
satisfied, or whatever, and whichever action leads to the greatest quantith is the right
action. Butthhe standard objection, running from Immanuel Kant (18 century) to John
Rawls (20 century) is that an action can be morally unacceptable even if it maximises
the aggregate good, because it might require the sacrifice of some particular individual’s
good. On Rawls’formulation, utilitarianism allows you to sacrifice the rights of some
individuals for the sake of others, so doesn’t take individuals seriously.
Deontology: Theories that emphasise non-reducible rights and duties are called
deontological. Kantian ethics and the Rawlsian theory of justice are the most famous.
Rawls: individuals have rights that may not be sacrificed no matter how much aggregate
social utility is thereby produced. Even if everyone living on a block would derive great
happiness by expropriating one house and turning the lot into a playground, it is
impermissible that they do so without the owner’s consent; otherwise they are violating
his right to his property (which is equal to their rights to theirs).An ethnic majority may
not oppress a minority even if the overall ensuing happiness outweighs the increased
misery of the minority (say the minority was docile enough not to mind too much, and
that the majority was so selfish as to derive great pleasure from the exploitation). 1. Objections to hedonic act-utilitarianism inaugurates a long history of backtracking.
Here are some important objections (by no means exhaustive):
(1) Incommensurability: Can different pains or pleasures be compared? Can one really
quantify over a wide range of different comforts and discomforts, pleasures and
displeasures, ecstasies and agonies, etc.? Can one assign quantities to experiences of mild
pain, mental anguish, general displeasure, indignation, all on the negative scale, and then
joy, contentment, exhilaration, appreciation, approval, all on the positive side, and each
one of these of different objects, and somehow be able to weigh these all against each
other? If not the entire project of utilitarianism would seem to collapse.
(2) Adaptive preferences – people can adjust the degree to which they are satisfied by
lowering their demands, but is this good? InAesop’s fable, the fox that couldn’t reach the
grapes says, Oh, they must have been sour anyway. People often think this way when
they can’t get a job, or are poor, or even when they are victims of injustice. Oppressed
minorities may internalise the negative views of them by the dominant group, and then be
more inclined to accept their disadvantaged lot. Women may come to think in a
patriarchal society that their place in the home is “natural”, virtuous, or commanded by
God. But insofar as they come to be satisfied with their oppression by not viewing it as
oppression, this internalisation will raise aggregate welfare levels. But surely that is
morally unsatisfactory. If the utilitarians’original intuitions were that a good ethical
theory will strive to satisfy as many people’s preferences as possible, it must be a serious
problem if many of those preferences are merely adaptive: people lowering their
expectations and accepting (‘preferring’) their lousy lot in life just because it seems that
they can’t get any better.
(3) Multiple values. Happiness ain’t everything, especially if understood hedonically as
merely pleasure and absence of pain. Dignity, truth, beauty, creativity, understanding, all
might be important values that are not reducible to pleasure. Genuineness of experience
might also be valuable in itself, or valuable to realise these other values. Nozick’s
experience machine: if you agree that you wouldn’t want to spend your life plugged into
a machine that gave you the illusion of having all sorts of wonderful experiences that
weren’t genuine (and hence the accompanying sensations but without the real
experiences), then you probably think that the genuineness of experience, and not just the
effect it has on your mental state, is valuable, especially if it is a good experience. That is
one of the reasons why use, or at any rate overuse, of recreational drugs is frequently held
in disdain. Beyond happiness, not only living in reality, but being creative and
productive, having real friends who like you for your real character traits, people who
admire you for your real achievements, and so on, is important; so not just pleasure but
how you get the pleasure.
(4) Justice (most important objection) - utilitarianism is blind to distribution. Consider: torturing an innocent to please sadists. Say torture produces x amt. of pain. So
x amount of pleasure + 1 util experienced by sadists witnessing the torture would tip the
balance in favour of doing it. There must be some number n of sadists who would derive
more than x satisfaction. (Suppose the video goes viral among the sadist community. Or
consider child porn watched by a “sufficient” number of paedophiles so that their
pleasure outweighs the harm done to the child).
Justice seems to imply fairness. Fairness seems to imply that everyone be considered
equally. Utilitarianism doesn’t care directly how pleasure and pain are distributed; it only
cares about maximising the aggregate.
Suppose you think equality is valuable. So a distribution of 10, 10, 10 in a society would
be better than 4, 9, 17 or 7, 10, 13. But util. seems to say that 4, 4, 23 is better than all
these, since the aggregate is higher. Note: you might allow that 10, 10, 15 is better than
10, 10, 10 since no one is made worse off, but think that 8, 12, 15 is worse than 10 10 10
because one person is made worse off. These questions take one to the heart of
distributive justice, but utilitarianism seems oblivious to the intrinsic importance of
(5) Overdemandingness: a theory that tells you to maximise the good in the overall state
of affairs of the world is an incredibly demanding theory, which might enjoin you to
abandon all your personal goals, attachments, projects, and values. More on this below.
Some students upon first learning about utilitarianism confuse it with egoism: that each
individual should seek to maximise her own pleasure or happiness or desire-fulfillment.
Utilitarianism is the very opposite: it enjoins one to maximise aggregate happiness in the
entire universe. In that sense it is a highly altruistic theory that even obliterates the
distinction between the supererogatory – going beyond the call of duty – and what
ordinary duty requires. This objection is that it is too altruistic, to the point of being
overdemanding. Your own family, friends, and projects are only worth their tiny
proportion of the happiness of the entire universe; their special significance to you counts
for nothing, except in that you my be in a better position to affect them.
Consequentialism: 2 and 3 are possibly answered by shifting from utilitarianism to
consequentialism, the view that an action is right if it maximises the good in general,
which need not be merely pleasure or happiness. Sophisticated versions allow that the
good can be constituted by different values: liberty, welfare, happiness, and even moral
goods such as equality. (You could have an egalitarian consequentialism that judges an
action to be right if it leads to greater equality in the overall state of affairs. You could
even be a rights-consequentialist: an action is right if it maximises respect for rights – see
Consequentialism: The term was coined byAnscombe as a term of opprobrium – it was
what was wrong with utilitarianism in her eyes. But then several philosophers with some
utilitarian inclinations but who were sensitive to the above longstanding objections to
utilitarianism took up consequentialism as a positive doctrine. From Bentham to Mill:
The early utilitarian tradition with Bentham was very robust – very reformist.
The generation after Bentham became much more apologetic and accommodating,
claiming that utilitarianism just explains most of our ordinary moral traditions, and
captures what’s best inAristotle, Jesus, etc.
Bentham had provocatively opined that it was better to be a pig satisfied than Socrates
dissatisfied. Along tradition of backtracking from this began in earnest with Bentham’s
successor, John Stuart Mill.
Backtrack 1: evaluative qualitative distinctions between pleasures: higher and lower
Mill took it all back: it is better to be Socrates dissatisfied than a pig satisfied. Why?
Because Socrates has higher pleasures that are worth more than the greatest amount of
Haydn and the Oyster: you are a soul, and an angel asks you to choose between two
options: be the great composer Haydn, who lived a normal 18 century human life span,
or be an oyster, but with a life span so long – say, 10,000, or 100,000 years – that the
pleasure it will contain will be greater than that experienced by Haydn.
If all pleasures are commensurable, such that only the intensity and duration vary, in
other words the only differences are quantitative, then there must be some duration long
enough that even the low intensity pleasure of the oyster will outweigh the short duration
‘high intensity’pleasure of Haydn.
Mill: Higher pleasures are infinitely more valuable than lower pleasures, as proved by the
fact that anyone who has experienced both will always prefer the higher to the lower.
Sometimes Mill speaks of the higher as intellectual. So, anyone who learned chess will
never go back to playing chequers because the former is more intellectually interesting.
But sometimes Mill speaks as if any two pleasures can be compared to see which is
‘higher’and which is ‘lower’.
‘Higher’here sounds like higher value. Mill says: Pleasures vary in quality too, not just
quantity. Yes, but ‘higher/lower’implies a distinction not just between qualities in the
non-evaluative sense, but qualities in the evaluative sense. It is not just saying that the
taste of chocolate is a different qualitative experience than listening to Schubert’s
Unfinished Symphony. It is saying that one is better, higher, than the other. But if that is
so, Mill can’t make that consistent with mere hedonic utilitarianism; he can’t make these
rankings consistent with saying that pleasure is the only thing of value, because he needs
some other value in order to call one kind of pleasure ‘higher’(i.e. more valuable) than
another kind of pleasure. Philosophers in other traditions – Kant,Aristotle – would say, Of course doing philosophy is more valuable than the pleasure of a pig wallowing in mud
– it’s using reason and intellect, and therefore has dignity, is ennobling, and so on. But
Mill can’t quite allow himself to say that because he’s still trying to be loyal to the dogma
of Bentham and Mill’s father, James Mill, that there is only a single value: pleasure.
2 backtracking: justice and equality
Mill also argued that there were good utilitarian grounds for justice and equality. But
even if his arguments are correct that aggregate happiness will be increased by following
justice and equality, is that the reason why one should be just? Is that the reason why I
shouldn’t steal an old lady’s purse – because doing so will lower aggregate happiness
Is the reason governments shouldn’t arbitrarily arrest anyone that doing so would make
everyone in the long term less happy, or is it not an injustice to the very people arrested,
plain and simple?
Mill thinks justice marks out the most important human interests, and thus the most
important parts of happiness. But the question remains: why not sacrifice some
individuals’happiness for the sake of others if you could maximise the aggregate that
way? Mill implies that doing so would not maximise happiness, but even if that “actuarial
consideration” (Rawls) were true, is that the right reason? Is it not simply that it is unfair
to do so, because it violates a person’s rights? (That is the deontological view).
Another conciliatory argument later utilitarians developed is that of declining marginal
utility. The more income, say, you give to one person, the less benefit they will derive
with it compared to the benefit to someone much worse off. So that is a reason to
distribute income and resources more equally.Again, even if true, is that the only reason
to do so? Deontologists think fairness and justice are reasons in themselves, not mere
instrumental strategies for maximising long-term happiness.
3 backtracking: preference satisfaction
Some utilitarians try to maximise preference satisfaction rather than happiness, to get
around problems of incommensurability. But what about external preferences? It’s fine if
everyone has preferences just about their own lives in ways that don’t affect others. But
people also have preferences about other people’s lives and property. And why should
racist, sexist, or selfish preferences count equally with others?
4 backtracking: consequentialism, not utilitarianism. Consequentialism can
accommodate more values than happiness or ‘utility’.
But act-consequentialism (that you should always do the act that maximises the good
overall) is just as vulnerable as act-utilitarianism to a massive objection: that the theory is
overdemanding (objection 5 above) and undermines personal goals, attachments, commitments, and values. On some formulations it is also self-defeating. So this leads to
the fifth major backtracking from the original Benthamite hedonic act-utilitarianism.
5 backtracking: rule-consequentialism, not act-consequentialism; or, if you still prefer
utilitarianism, rule-utilitarianism, not act-utilitarianism.
Act and rule consequentialism:
Different theories have different approaches to rules. Even within a theory such as
consequentialism, there are different approaches to rules.
2. Act-consequentialism (act-c.):An act is morally right if, of all possible acts available
to the agent at the time, it is the act which leads to the best consequences overall (in the
3. Objection to act-c.: If everyone tried to follow act-consequentialism they would
actually end up making everyone worse off. Reason: If they constantly had to decide
which act would promote the best consequences, they would have to be willing at every
moment to abandon personal commitments, attachments to loved ones, friends,
compatriots, personal projects, and rules promoting sincerity, loyalty, devotion, and so
forth. But abandoning or undermining these would make us all worse off, since these are
among the most important goods that we have. Note: the objection is not: if everyone
followed act-c. we would all be worse off. Recall:Act-c. says: do the action that will
maximise the good. So it is logically impossible that everyone be worse off if everyone is
always actually maximising the good. So if it’s true that undermining sincerity and
commitments, and abandoning personal projects and attachments, would at some point
actually tip the balance toward making things worse, then act-c. would say not to do it.
That’s why the objection stresses that everyone would be worse only if everyone tried to
follow act-c., not if they actually did follow it.
Sometimes the objection is put in the form of a claim that act-c. is self-defeating. The
point can be put this way: Since if everyone tries to follow act-c. they will actually make
things worse, maybe the best way to promote the best consequences is if everyone just
follows their ordinary morality. But then the theory is pretty useless and even self-
defeating, since it will now amount to the peculiar theory: to realise the aims of act-c.,
ignore act-c. and just follow your common moral views (don’t lie, don’t cheat, be loyal
to friends and family, pursue your personal projects, etc.).
4. Rule consequentialism (rule-c.): In face of the above objection, some theorists have
tried to save what seems to them the element of truth in consequentialism – that acting
morally must have something to do with producing good consequences – while
combining it with the agent following rules (which the deontologist opponents make
central). They thus propose rule-consequentialism (as an alternative to act-
consequentialism): an act is right if it conforms to rules the general compliance with
which would lead to the best consequences. So you choose rules because if everyone
followed them it would lead to the best consequences. It’s still consequentialism because the justification of the rules is in terms of the good consequences they would promote.
And the theory is still trying to maximise this good. But it does so now by saying that
everyone should follow certain rules.
5. Objection to rule-c.: Some object that rule-c. collapses into act-c.:Assume a set of
rules: be sincere, be loyal to friends and loved ones, pursue your personal projects with
devotion, as well as do not kill, steal, etc.According to the definition of rule-c., they are
selected because general compliance with them would promote the best consequences.
But if the rules are any good, that is, if they are reasonable and really have a chance of
promoting the best consequences, they will not be crude but instead incorporate
exceptions into the rules. For instance, do not kill admits of the exception of self-defense.
Do not lie, on a reasonable interpretation (i.e. not Kant’s!), presumably admits of the
exception: unless to save an innocent life. So originally crude rules can be refined by
modifications that incorporate exceptions. But then any time some actA would promote
better consequences by violating a rule than any alternative act that complied with the
rule, one should presumably modify the original rule to incorporate this exception. But
now we are back to the act-consequentialist view that says you should evaluate each
situation independently of the rules and decide what to do on the basis of which act will
promote the best consequences. In other words, each rule becomes modified or refined to
such an extent – with exceptions to exceptions to exceptions, etc. - that the entire corpus
of rules becomes co-extensive with act-c. Finally, we can replace this elaborate corpus of
rules with the single rule of act-c.: perform the act that will promote the best
consequences. (After all, that’s a rule too, so act-c. is like a on